Vehicle navigation strategy based on behavior fusion

Abstract

In this paper, a new navigation strategy based on the fusion of various behaviors to enable mobile vehicles to navigate in an unknown environment is described. The aim of this research is to fuse the two independent and sometimes conflicting behaviors: obstacle avoidance and goal seeking, such that the vehicle efficiently performs obstacle avoidance while seeking it's goal. The balance between these two behaviors is achieved by combining the control actions from the goal seeker and the obstacle avoidor through evaluating the goal vector magnitude, the minimum distance detected by the ultrasonic sensors, and the distance to the obstacle in the direction of the goal vector. Furthermore, an environment evaluator is used to enhance the adaptability of the navigator by tuning the universe of discourse of the sensor space. The new navigation strategy has been verified to be efficient in an indoor virtual environment.